1. Technical Field
Present invention embodiments relate to database replication systems, and more specifically, to idling replication of a specified database object during database replication.
2. Discussion of the Related Art
Data replication services store data from a source database in databases of separate query machines thereby providing the ability to run queries with up to date information but without burdening the source database.
When a data replication service experiences high latency, the information on the separate query machine may only be current as of the latency period. For instance, if latency is one hour, results of a query would only have information from an hour ago. If only a particular table or a particular set of tables need to be queried, it might be faster to load data directly from the table in the source database into the particular table or the particular set of tables on the separate query machine so that the query can be run on the separate query machine with up to date data.
With replication for a particular table on the separate query machine active, however, the particular table cannot be loaded on the separate query machine until replication operations on the particular table are completed and/or stopped, i.e., until replication of the particular table is idled. Otherwise, errors will occur. In order to idle replication of the particular table, the entire replication needs to stop for all tables of the replicated database, thereby requiring a restart of the entire replication process.